Martensitic steels are widely used in many areas such as automotive, mining, and agriculture mostly thanks to their thermal loading ability property. On the other hand, these special steels exhibit extreme tool wear tendency and low surface quality which can be associated with abrasive resistance. This situation makes this steel hard-to-cut and requires further investigation with several approaches. Sustainable machining environments are highly effective as modern strategies to improve the machinability index. Also, machine learning models have pivotal role on decreasing the total consumption in the way of lean manufacturing. In the light of above-mentioned information, this work focuses on the machining performances and optimization of dry, flood, and MQL conditions during the milling of Hardox 400 martensitic stainless steel. A novel approach was applied with using several cutting environments and machine learning models to enhance machinability of Hardox which is an industrially important material. Results were analyzed with different machine learning models using heat map and decision trees. Seemingly, cutting fluid assistance in the milling of Hardox steel is critical where flood and MQL provided a considerable effect on the tool wear for reducing it under some level. Also, this technology was found useful in determining the best conditions of machinability in terms of surface roughness, chip morphology, energy consumption, and cutting temperatures. Machine learning models provided hopeful results in analyzing the correlations between parameters used in the model. In machine learning, the heat map being close to 1 and the MSE and MAE values being close to 0 indicated that the model was suitable. This study is expected to observe the contributions of different types of cutting environments to the machinability criteria during milling of industrially important materials.
Read full abstract